| |
|
There's a difference between CG as effects work - stuff which would previous have been done optically, like mattes and so on - and CG in the creation of characters and monsters. The first (and some extensions like duplications for crowd scenes in lower budget pictures, as with the massive poltical rallies in Richard III - shot in a school gymnasium with about twenty extras) is pretty much inevitable, and shouldn't really be a problem. Although it can allow Matrix-like effects into movies which probably shouldn't have them, which is a drag, most of the time you won't notice.
The second hit the bigtime pretty much with T2: the creation of fully-rendered artefacts which can interact with live actors. Or not interact, as in the case of Sky Captain. The trouble with this stuff is that however good the physics engine is or however many polygons you use, ultimately - for the moment at least - the modelling isn't good enough - at usable levels of pricing - to generate something which is totally convincing. And even that's not a terrible disaster - movies haven't got to look totally convincing, they just have to look good enough that you go 'ooooooh' rather than 'yeah, whatever'. The real problem is that they all look wrong in the same way. That slightly-too-smooth finish, the perfect synchronisation of the footsteps of a marching army - even the statistically accurate out-of-synch footsteps of more recent efforts; these are all the things which trip our CG bullshit sensors, along with cues you almost couldn't identify, like eyelines, dust in the air, depth of field, blurring, angle of light... We're machines for spotting things like that as a consequence of our evolution.
All of which is a long way of saying that you probably could make a superhero movie without obnoxious CG, but you might end up with more obnoxious animatronics, or a vastly-reduced field of villains and life-or-death situations. |
|
|
